There are few more profound and, unfortunately, permanent losses than those seen in spinal cord injury at the neck, (SCI). The person may lose all sensation and movement below the neck, while gaining urinary and faecal incontinence. With such profound loss of sensory input and motor output the person is faced with a 'Cartesian nightmare,' disembodied but thinking. This paper explores some aspects of sensory and motoric imagination found after SCI.

Imaginings must be distinguished from hallucinations and immediate phantoms: it is not uncommon after acute SCI for people to feel their legs to be in unusual or impossible positions: these thoughts usually fade. Pain does not invariably accompany SCI but does develop and become chronic in 60-65% of those with SCI. Usually this pain is felt below the level of the injury in an insentient region. This deafferentation 'phantom' pain is not considered imagined since it is not a cognitively elaborated experience. Perhaps surprisingly this deafferentation pain can be considered a 'friend' since it allows the perception of a body otherwise insentient and allows the subject to believe, or imagine, a continued wholeness of body.

People with SCI vary in the extent to which they claim to have a normal experience of the immersed imagination we all might elaborate during watching sport: we normally identify with the participants, whether they are professionals or children in a garden. Some with SCI suggest they feel as fully involved as pre-injury while others feel unable to enter the experience of others in this and feel a spectator; motor empathy seems, like so much in those with SCI, to be varied and idiosyncratic.

In dreams people with SCI seem initially to be whole and walk normally. As the years after injury unfold people become more disabled in dream. One person says that she takes her wheelchair on her back for security during dreams, while another signals the end of the dream by getting back in her wheelchair.

But perhaps the most intriguing imagined experience after spinal cord injury is sensory. Several people have suggested that they 'feel' touch on areas of skin. Subjects say that they feel touch when they see their legs touched, in areas of skin without cutaneous sensation. There is no touch if they do not see contact. This seems to be a strong illusion, (and one subject suggested the same occurred during sex), cognitively elaborated, top- down, but which becomes almost an automatic 'seen touch.'

Lastly motor imagination can be assessed in SCI with fMRI imaging. Similar areas of the brain are activated during de-efferented, imagined, movements in those with SCI to normal subjects actually making movements, (Shoham et al, Nature, 2001, 413: 793.).

These elaborated sensations and motoric imaginations, 'seen touch' and low chronic pain and motor ideation, may be important to the person with SCI's maintaining an embodied existence and connection with the body they can of course see but not move or feel.

For Descartes, the imagination was simply our faculty for producing a mental image. He distinguished it from the understanding by noting that while the notion of a thousand sided figure was comprehensible÷that is, was sufficiently clear and distinct to be differentiated from a thousand and one sided figure÷it could not be clearly pictured in our mind. The representation of its sides exceeded our powers of imagaination. This view of the imagination as our ability to produce a mental image fails, however, to distinguish it from remembering. Let us say that I see an object and then I close my eyes, maintaining the image of the object. Is this imagining or short term memory? What about the case, when an hour later I recall this image. Am I imagining or remembering it? I am going to assert that imagination, as distinct from memory, implies something more than the ability to produce a mental image. It involves, as Sartre pointed out, a certain attitude towards this image. Engaging in it, we deny its reality. In Sartre's words, imagination "carries within it a double negation; first, it is the nihilation of the world (since the world is not offering the imagined object as an actual object of perception), secondly, the nihilation of the object of the image (it is posited as not actual) ..." (BN, p. 62). Imagination, then, represents the imagined as nonactual.

In this paper, I am going to examine this sense of the imagination. Specifically, I will explore how we make the thesis of the nonreality of what we imagine. I will then note the implications of this for machine intelligence. My conclusion will be that imagination involves our relations to others. It is through them that we both advance and withdraw the thesis of the reality of an object or state of affairs. Thus, we take as real what we all agree to be the correlate of our separate experiences. We negate this thesis, i.e., take as nonreal, whatever fails this test. The nonreal is what only one person experiences, i.e., what the present intersubjective context does not verify. This nonreal, however, is not nothing. It is that which would be real were a corresponding intersubjective agreement available. It is precisely this thought that distinguishes imagination in the Sartrean sense from the simple maintaining of a mental image. Crucial, then, for the imagination is our experience of differing intersubjective contexts and agreements. These, in fact, supply our imagination its materials. They also provide the alternate contexts through which we can see our presently experienced ãrealä world as not offering the imagined alternative. This implies that to give a machine intelligence the possibility of imagining in the Sartrean sense would be to give it a corresponding experience of "intersubjectivity." It would be to link it to a community of other machines, each of would inform the others of their activities, thus, stocking their abilities to imagine alternatives to their present imputs and procedures.

An integral part of phenomenological method is imagery. Imagery is involved in understanding descriptions of things or events not originally present to oneself, in "simple" abstraction (distinguishing dependent, i.e. abstract parts from one another or from a whole), in generalizing abstraction, formalizing abstraction, and "eidetic intuition". But imaginary acts are only indirectly observable. Hence the practicability of phenomenological method and its claim to objectivity ought to be reconsidered with regard to its dependence on imagination.

The case of auditory imagery will serve to outline some of the problems arising within this area and the practicability of phenomenology. E.g., it is not in any doubt that loudness is a necessary dimension of every consciously perceived sound. But what about loudness of imaginary sounds? There seems to prevail the opinion that there is a pendant to perceived loudness, but not "all the time". (Cf. M. J. Intons-Peterson (1992), 'Components of Auditory Imagery'. In D. Reisberg (ed.), Auditory Imagery. Hillsdale, NJ: Lawrence Erlbaum Associates.) But this still can be contested since it is arguable that not every subject is able to imagine loudness or to focus on imagined loudness, or even that there is no imaginary loudness at all.

I shall argue that subjects, at least implicitely, apprehend imaginary loudness whenever they imagine auditory contents (e.g., inner speech). Among the reasons supporting this claim are the following arguments: (1) When I "take away" the loudness of an imagined sound this (quasi-)sound ceases with its (quasi-)loudness: Exactly as it is impossible in auditory perception to perceive a sound without loudness, it is, it seems to me, impossible to imagine a sound without quasi-loudness. (Though I do not claim that auditory imagery is exclusively determined by auditory perception. This is rather improbable since auditory perception too is not determined by auditory information exclusively.) (2) Imagine a rhythm that is well articulated, for example the rhythm of "Old MacDonald" as a sequence of tones of one single pitch played by a piano. Do you "hear" accents, is there a change of weak and strong tones? If there is then there is quasi-loudness since there is hardly any other dimension capable of giving the "impression" of (imagined) accents in ths case. Put generally, where there are accents, there are differences of quasi-loudness, and thus there is quasi-loudness. (There will be presented several observations, crosschecks and further arguments to buttress those suggestions.) - The reason why (many?) people think there is no quasi-loudness might be that it is very difficult for them to focus on certain abstract dimensions (e.g., loudness) as part of an imaginary content. But the fact that they are not aware of it does not imply that it is not there. (Most listeners cannot specify the perceptual dimensions of perceived sounds either.) This is where phenomenological "practice" comes in.

Finally I will be considering a possible remedy against the "subjectivism" (and hence the missing validity or objectivity) of phenomenology, namely doing phenomenology cooperatively, taking as an example agreement procedures in phonetics. (There are some striking similarities in problems faced by phoneticians who wish to obtain maximally objective transcriptions of auditory data and philosophers who wish to obtain maximally objective descriptions of mental operations and contents.)

What are the evolutionary and neurocognitive conditions which enable the temporary suspension of belief in a background reality to become engrossed in the imaginary contents of the narrative frame, allowing a second temporal sequence to unfold? M. Donald's proposal of the development of mimetic skills in the evolution of conscious episodic memory is one direction of possible explanation. If Sartre is correct, however, the awareness of nothingness in the upsurge of consciousness is completely contingent - a dialectical process that proceeds by negation - and therefore could not be selected for in the evolution of higher cognitive functions. Yet, it is the framing function, the ability to temporarily suspend belief in background reality to become more imaginatively engrossed in the narrative frame - also seen as a dialectical process between engrossment and reflective distance - which is lost in early psychosis. This suggests a neurocognitive foundation that is disrupted by the psychotic changes. The loss of a framing function is corollary to the observation made by the phenomenological psychiatrists Blankenburg and Stanghellini that there is an "involuntary" or imposed phenomenological reduction in early schizophrenic psychosis. After all, if one is no longer to flexibly shift from the phenomenologically reduced contents of consciousness as foreground back to the suspended reality of everyday consciousness as background, the ability to narratively frame events as imaginary has also been lost. For Husserl, imagination involves the neutrality modification (or temporary suspension) of the event's character as "being," i.e., an ability to disbelieve. In the presentation, I will present the phenomenological theories of imagination in early psychosis, and will propose the neurocognitive conditions of the framing function and its disruption in different disorders (in schizophrenia, as well as, in a different way, the trauma-related disorders). I will also present examples of the strategies of patients with schizophrenia in their attempt to recover the frame with their spared cognitive abilities.

Emotions are no longer considered to be mere noncognitive bodily responses or impediments to rationality. They are now widely recognized as central to biological cognition, and as bearing intentionality. Less widely discussed is the role of imagination in connecting emotion with consciousness. I will explore this topic from the perspective of embodied phenomenology. I will argue that emotional experience is a variety of imaginative cognition, conditioned by the body scheme. Imaginal models are qualitative presentations or enactments of lived experience. They are like maps of our own forms experience - patterns of activation within the body scheme. Because of the embodied nature of consciousness, these activations inevitably evoke an emotional value for those who experience them. Drawing on Gallagher (1996) and Damasio (1999), I will develop this approach by contrasting it with two other contemporary views of emotion, those of Nussbaum (2001) and Ellis & Newton (2000). Both lack a phenomenological analysis of imagination.

I. Imaginal Models
An image is only imagined when it is taken up within the experience of an embodied subject. Philosophers and cognitive scientists have used the term 'representation' in several ways. For analytic philosophy of mind, a representation is a syntactically defined data structure in a physical symbol system. It is a sub-personal (or third-person) element that need not be accessible to consciousness, even in principle. Cognitive neuroscientists use 'representation' in a similar way to mean a trace in the brain. For the phenomenological tradition, however, a representation is an intentional object or appearance of thought that can in principle become conscious. It is a noematic structure that is accessible only from a first-person perspective. This distinction also holds between an image and an imagining. Imagination is the creation of imaginal models for consciousness in accord with the lived structures of the body scheme.

Models are representations, but not statements or propositions. They are neither true nor false, but like maps, they can 'fit' some area of the world or zone of experience more or less well. The structural fit of a model to a domain is a matter of the degree of resemblance between them, under an interpretation. What counts as fit is relative to the map makers and map users. No map is a complete map - it doesn't exhaust what it represents and other maps are always possible. A map is a structural representation of some salient features of the terrain, chosen according to the particular purposes of the mappers. All models must be interpreted from within a perspective, as all maps must have a key. An imaginal model is a presentation or enactment of the qualities of an embodied experience. Imaginal models, then, are qualitative maps used by conscious organisms. They are presentations of 'what it is like.'

Recent work by Anotonio Damasio (1999) helps illuminate the affective dimension of imaginal models. He defines emotions as complex, stereotyped patterns of response in the body scheme or 'internal milieu' - the visceral, vestibular, and musculoskeletal systems as well as numerous brain circuits. According to Damasio, emotions can be induced in two ways. First, when an organism perceives certain objects or situations, and second, when an organism conjures up certain images in consciousness. Damasio argues that although the physiological machinery for emotion is largely preset (i.e., the body scheme), the inducers of emotion are not. The specific ranges of imagery that become inducers for classes of emotions are variable and external to the body scheme. But because of the embodied nature of consciousness, imagination inevitably carries an emotional tinge: 'Another important consequence of the pervasiveness of emotions is that virtually every image, actually perceived or recalled, is accompanied by some reaction from the apparatus of emotion.' Thus, the body scheme invests intentional objects with emotional values.

II. Nussbaum vs. Ellis & Newton on Emotions
Nussbaum (2001) defines emotions exclusively in terms their evaluative content. Thus, she gives an essentially functional account predicated on the denial of type identity between emotions and physiology. Her definition is designed to make both consciousness and imagination incidental to emotion. I argue that this approach will run into the same problems that traditional computationalism does in understanding the meaning of representational states for the organism.

Closer to the present perspective, Ellis & Newton (2000) hold emotions to be properties of self-organizing systems, the aim of which is to maintain organizational balance within the self-organizing process. They also claim that emotions are the key to distinguishing conscious from nonconscious information processing. But their account lacks a theory of imagination. Instead, they simply distinguish between those self-organizing processes that are representational and those that are not. This is inadequate. Emotions need to be situated within a phenomenological account of imagination that connects them, on the one hand, with embodiment, and on the other hand, with consciousness.

Despite Husserl's well-known and often quoted statement that fiction is "das Lebenselement der PhŠnomenologie," that is, the vital element that makes up the very life of phenomenology, his account of phantasy has been repeatedly characterized as unsatisfactory. The most common accusation is that Husserl, although he rejects the simple empiricist view of imagination as a 'faint copy' of sensation, likewise degrades phantasy to a poor imitation of perception by describing it as quasi-perception. Further, Husserl has been criticized for remaining indebted to empiricism by claiming that phantasies are produced by a specific apprehension of sensory contents. Finally, while Husserl is often credited for ascribing some methodological merit to phantasy in his notion of eidetic variation, this 'merit' - the possibility to apprehend an essence by freely varying a noema in phantasy - is often itself perceived as one of the most questionable elements of the phenomenological method. In the light of these objections, Husserl indeed appears to be "yet one more in the tradition of writers whose words do not match his deeds in the affairs of the imagination."

However, the enigma of phantasy as the Lebenselement of phenomenology remains. It is quite remarkable that Husserl discusses phantasy not only in the manuscripts explicitly engaged in its description (published as Hua XXIII by Eduard Marbach in 1980), but in almost all his philosophical works spanning from the Logical Investigations (1900/01) to the posthumously published Experience and Judgment (1939). Perhaps this puzzling circumstance can provide us with a possible point of entry to the problem. It suggests that only if the whole range of Husserl's investigations is taken into account, can the question of the role of phantasy in Husserlian phenomenology be adequately addressed.

I take up the question of phantasy's vital function and propose that it occupies a critical systematic place in Husserl's work. However, in order to disclose this place it is, first, necessary to criticize the account of phantasy that Husserl submits in his early writings. Simply put, it consists of an application of the 'content-apprehension-schema' to phantasy, which is modeled after his theory of picture-consciousness [Bildbewu§tsein]. It assumes the occurrence of phantasmata turned into phantasies by a specific mode of apprehension-a conception which, in my opinion, indeed justifies the charge of empiricism.

In the second section of this paper, I draw attention to an alternative notion which Husserl develops in his writings from 1905 onwards. Pressed by the problems raised by his early conception of phantasy, he eventually recognizes the insufficiency of his 'content-apprehension-schema' and admits its latent representationalism. To the contrary, he now claims that "consciousness consists of consciousness through and through, and even sensation as well as phantasma is consciousness."

As I propose in the third and final part of my paper, Husserl thereby takes the crucial step towards a transcendental phenomenology of imagination. By means of this transcendental account, he is able to reveal the constitutive function of phantasy in the realm of possibilities and a priori cognition. Moreover, he also refutes the supremacy of perception hidden in his early approach and overcomes his own empiricist legacy in order to give way to a truly 'first philosophy of phantasy.'

Recent neuroscientific literature on imagination is replete with references to representations in the brain. Classic representationalism andfunctionalism combine in much philosophy of mind to provide an interpretation of such references. In this paper, I will develop an alternative to the Cartesian picture that these two approaches have encouraged.

Classic representationalism understands psychological states typically to possess intentional or quasi-semantical contents. Such states are referred to in recent (analytic) philosophy of mind as "mental representations." The first and central task for this paper is to provide an alternative interpretation of "representation" as it occurs in much neuroscientific literature. So interpreted, "representation" does not carry any implications about the possession of quasi-semantical contents. Consequently, according to the arguments of this paper, neuroscientific uses of "representation" need not imply any commitment to classic representationalism.

A functionalist approach to mental kinds takes them to be complex causal states that are in fact neural states. A second task is to examine some issues about the kinds of term or classification "imagining" and "imagination" are. Can they be given a purely causal or purely individualistic interpretation? Or do they carry with them normative implications and reply on standards or practices in the society?

I argue that at least some neuroscientific representations are not "mental representations" understood as possessing quasi-semantical contents with satisfaction conditions (and so matching or failing to match the world). Rather, they are token-realizations, which are just examples. The difference between these two is illustrated by the difference between describing a noise and repeating it. A description will have a content that is true or false. In contrast, if I simply repeat a noise, I do not say anything about it; rather, I simply exhibit it or give an example of it. The latter may remind us of an ancient model of representation, present in philosophy at least since Aristotle, and, I argue, alive and well in today's neuroscience. (Isomorphism becomes essential to the account, for it allows us to avoid the Aristotelian thesis that sensory qualities and essences literally get realized in the mind.)

I contrast contemporary uses of "imagine" and "imagination" with Aristotelian and related Cartesian uses. Unlike these earlier notions, these ordinary terms do not connote a natural kind. Rather there is a wide range of uses from "She has a wild imagination," to "I imagine that most people find the play boring." (Cf. Wittgenstein, Investigations, Part II, on "I am afraid.") One aim is to undercut any simple identification of imagining with neural processes. Another is to examine some of the ways in which what counts as imagining is normatively constrained by standards present in one's society.

The function of the imagination in the aesthetic realm is different. Kant considers two aesthetic phenomena: the beautiful and the sublime. When the mind lingers in disinterested contemplation of a beautiful object, the imagination is engaged in free play. The beautiful also reflects the harmonization of the imagination and understanding. The aesthetic pleasure of the beautiful points to two important architectonic aspects of human knowledge: the possibility of the harmonization and integration of the powers of the mind and the suitability of nature to our human faculties. The experience of the sublime leads to even further modification of the imagination. With the sublime, reason demands of the imagination that it present an image of the immensely large, the formless, the incommensurable. The imagination is unequal to the task. Kant himself calls the sublime state a "a subjective movement of the imagination by which it does violence to the internal sense" (CJ, 108), such as is needed for cognitive experience. Moreover, the sublime state is characterized not simply by an unmixed feeling of pleasure, as is the case with the beautiful. Rather, the sublime evokes both pleasure and pain. The pleasure arises from the exaltation of reason and its supersensible destination; that is, its ability to rise above sensible nature. The pain comes from the failure and submission of the imagination. Sublime consciousness is accordingly one of agitation and disorder because there is no simple union of the faculties.

In his Anthropology Kant catalogues a number of mental illnesses and deficiencies that are analogous to the sublime state of mind and the disorder of the imagination. One such malady (being crotchety or quirky) is the "perversion of inner sense" that "borders on insanity [Wahnsinn]." Under mental illnesses Kant lists rapture (raptus), a "sudden change of mood" (Anthropology, 83). He also registers a kind of delirium or derangement of the mind called Amentia (Unsinnigkeit), which is "the inability to bring one's ideas into even the coherence that is necessary to make experience possible" (83). Finally, in the General Remark of the Third Critique, which immediately follows the analytic of the sublime, Kant likens the sublime sentiment to the moral (or political) feeling of enthusiasm. Enthusiasm, Kant says, "is comparable to delirium" and is an affection in which the imagination is unbridled (CJ, 128). Moreover, enthusiasm is "a transitory accident to which the healthiest understanding is liable to become at times the victim" (128).

The "outrage on the imagination" (CJ, 91) brought about by the sublime can be traced to the modification in temporal functioning that the imagination undergoes. In cognition the imagination, as in all synthesis, is governed by time (A97), which is sequentially or successively ordered (A33). With the sublime, the imagination is called upon to produce an image of the absolutely great in a single moment, "at one glance [in einen Augenblick]" (CJ 107), thereby disordering the functioning of the imagination.

In this paper I argue that the disordered state of the imagination in the face of the sublime is in fact a delirious condition that Kant cannot contain within the merely secondary aesthetic phenomena of the sublime. The disorder of the imagination threatens to infect even the cognitive faculties of knowledge inasmuch as the harmonization of the faculties -- that is their being well-ordered -- is according to Kant the condition of any knowledge whatsoever (cf., e.g., CJ, 60). In short, if any individual faculty (such as cognition or understanding) can function only if the distinct faculties themselves work together, we do indeed have a disordering of all the senses when the imagination itself is diagnosed as delirious or disordered. I will suggest that the incommensurability of the Kantian sublime can be usefully viewed through Freudian lenses. The complexity of the sublime state of mind can be valuably analyzed in terms of repression and the unconscious. In short the Kantian notions of delirium and rapture that are linked to the psychological shock of the sublime and the disorder of the imagination are given a new valence in the lexicon of psychoanalysis under the names of anxiety and neurosis.

The lived body's experience and the experience of its intentionality towards the world is unitary thanks to the body schema, which is not a sort of map or static representation of the body's parts, but the dynamic integration of them in the organism's projects. The body schema is posture, a system of motor functions, which is orientated to a certain present or possible task. Intentionality, for Merleau-Ponty, is enacted by body motility. This intentionality, in Merleau-Ponty«s view, becomes Fungierende IntentionalitŠt, the true bodily intentionality, which goes beyond the bodily limits. The motor intentionality, which permeates Merleau-Ponty«s holistic phenomenology, comes from the body's unfolding into the world; it does not objectify, but it is the source of all the representations.

Imagination and emotion are not states but intentionalities. The motor inentionality of imagination does not need to represent the object to itself as a matter of fact but emotionally. Merleau-Ponty understands it through the imaginary sphere of the body, through the depth of the flesh, as the true Stiftung of the Being. By virtue of the fact that reason and imagination are originated in the lived body, Merleau-Ponty overcomes poststructuralist reductionism which is centred on the interrelation between disembodied signs.

Finally I apply the Merleau-Pontian phenomenology of imagination to an understanding of the contemporary iconic turn. The fast passing of images in our civilization nullifies their meaning, that is to say, the motor intentionality from which they arise. The conversion of a signal into bits, the digital image has broken with any physically discernible relation between object, percept and the imaginary. The body schema, the natural cultural root of images has fallen into oblivion. Merleau-Ponty reminds us constantly that the body or the incarnated consciousness is the ontological basis. He compares the body with the work of art which communicates ideas with colors and sounds by integrating phenomenona in a field. Art and schemata of the imagination arise from the experience of embodiment, but this experience is plural. The bodily schema is universal, but it is so in a diversity of situations, because it is a style.